DEPARTMENT OF GENETICS. 109 



tion and a lower specific electrical conductivity than are those of either of 

 the parent forms. The ratios of specific electrical conductivity to freezing- 

 point depression indicate that the hybrid absorbs relatively larger quantities 

 of electrolytes than either of the parent forms. The tissue fluids of the 

 hybrid are characterized by a lower acidity than are those of the Egyptian 

 parent, but by a higher acidity than those of the Upland parent. With 

 respect to this sap property, the heterozygous individual is, therefore, inter- 

 mediate between the two parent forms. 



A detailed discussion of the foregoing results is nearly ready for publication. 



These investigations have been continued during the present year, by a 

 more detailed comparison between the Egyptian and Upland types, involving 

 several newly imported Egyptian varieties. The F2 generation of hybrids 

 has also been investigated. These studies occupied the attention of Dr. 

 Harris, with the cooperation of Messrs. Arnold H. Johnson, Robert D. Evans, 

 and A. T. Valentine, during several weeks of the summer. 



Mirabilis. — Dr. John Belling has been working on the "multimutating" 

 genes of the four o'clocks, testing the hypothesis that the genes in the chromo- 

 somes have a number of possible allelomorphs to any of which they may 

 eventually mutate. 



"In the homozygous tricolor Mirabilis two pairs of multimutating genes 

 are found. The mutations are from yellow-striped to yellow, and from red- 

 striped to red, but the red shows only in presence of the gene for yellow. 

 The numbers of mutant progeny from tricolor parents were 6 per cent of the 

 yellow and 4 per cent of the red in 458 offspring. Totaling all cases in which 

 one of a pair of genes for yellow striping could mutate to a gene for yellow 

 there were 4 per cent of mutants out of 820 progeny; and in the similar cases of 

 red striping, 5 per cent of mutants in 1,063 progeny. In the parents of one 

 sibship, the gene for yellow striping was heterozygous, and in its progeny 

 there were 3 per cent of mutants out of 112 plants, where only about half as 

 many are expected as from a homoygous parent. A distinct mutant (prob- 

 ably chromosomal) has been found in tricolor progeny, of which the homozy- 

 gous condition seems lethal, and the heterozygotes occur in less than half the 

 normal number, 109 seeds giving a quarter of normals and less than a quarter 

 of heterozygotes, with only one recognizable mutant homogote." 



Genetic Basis of Animal Behavior. 



Heredity of behavior in dogs. — Considerable time was devoted during the 

 year under review to getting a quantitative expression of the behavior of 

 dogs as a preliminary to a study of the genetic elements in such behavior. 

 The work met with wholly unanticipated difficulties. These are described by 

 the experimenter (Dr. E. C. MacDowell) as follows: 



" (a) Discrimination apparatus. — The difference in the adaptability of two 

 litters of dachshunds in the preliminary training in the discrimination appa- 

 ratus was noted in the Year Book of 1921, pages 129-130. The training of 

 these two litters was continued for 6 months. For the first 3 months the 

 buzzers were used as signals; for the last 3 months the lights were the signals. 

 With the buzzers, there seemed to be a fair degree of learning shown by each 

 set, so that on the last 7 days of the training with the sound signals, no dog 

 made less than 8 correct trials out of the 10 per day. The difference between 

 the two litters in the preliminary training did not appear in their ability to 

 form sound associations. In the training with the light signals there are 



