DEPARTMENT OF EXPERIMENTAL EVOEUTION. 89 



ducing organisms there is a constant interchange of materials from one 

 stream to another. As one sometimes sees a muddy red stream emptying into 

 a clear river and can trace the color for miles, so in the mingling of germ- 

 plasms in sexual reproduction a single character may be traced through gen- 

 erations. In most organisms the mingling of germ-plasms is immediately 

 followed by the development of a new soma; and this is fortunate for the 

 biologist, since there is the least delay between making his mixture and learn- 

 ing the result of the mixture as revealed by the somas. In the new soma the 

 germ-plasm remains for a time in an immature condition, incapable of further 

 mixing. The usual period required for ripening is a year — and this fact lim- 

 its the speed of our progress. But in certain species generations succeed each 

 other with great rapidity; thus the vinegar-fly takes only 10 days betweerT 

 generations. This material has proved very valuable in the hands of Dr. 

 Lutz, who has reared over 40 generations in the last 2 years. 



From the study of mingled germ-plasms it appears that the same laws hold 

 whether the components be very unlike or differ in only a single character ; 

 for we have studied the effect of crossing distinct species and varieties, and 

 the behavior of individual differences in ordinary matings. The study of 

 hybridization is of prime importance for evolution, for it appears to be a com- 

 mon rule that a character that has arisen in one species of a genus is by 

 hybridization engrafted on to some of the individuals of various other species, 

 thus multiplying the number of species. This has been worked out by Dr. 

 Shull, of this Station, for the common shepherd's purse, Bursa. The same 

 thing is being worked out by Dr. Ezra Brainerd, of Middlebury, Vermont, in 

 the violets. And the lady-bird bettles (coccinellids) that Mr. Roswell Hill 

 Johnson has been working with since his connection with this Station prove 

 the same thing to be true for a large group of insects. These results could 

 hardly have been achieved, however, without a thorough knowledge of the 

 laws of inheritance of characteristics, and this Station has sought since its in- 

 ception to contribute to the determination of these laws. To this end experi- 

 ments have been made on crossing poultry, canaries, cats, sheep, and goats, 

 insects of several species, and many species of plants, and we have had asso- 

 ciated with us Dr. Castle, of Harvard University, who is working on mice, 

 rats, guinea-pigs, and rabbits. The general outcome of this work is the accu- 

 mulation of a lot of pedigree data— much of it quantitative — the like of which 

 exists, I venture to think, nowhere else in the world. These data have as yet 

 been only partially worked up,* but in general they confirm the fundamental 

 importance of Mendel's law ; they support, in many cases, the theory of fac- 

 tors by the union of which alone visible qualities appear and they have led 

 to a general theory of "dominance" in Mendelian inheritance according to 

 which a quality present in a higher degree dominates over the lower degree 

 or the entire absence of the quality. 



* Carnegie Institution of Washington Publications Nos. 23, 24, 49, 52, 70, 8r, 95. 



