444 EXPERIMENT STATION RECORD. 



nient of leaf blades due to torsion of the leaf sheath, not to a departure from 

 the usual alternate insertion of the leaf sheaths of gras-ses; (3) the develop- 

 ment of the silks within the leaf sheath instead of in a position exposing them 

 to the drying inllnence of sun and wind; and (4) a waxy endosperm differing 

 from both the horny and starchy endosperms of other corn in texture and optical 

 properties. The first 3 of these qualities indicate drought resistance. Analysis 

 showed the grain to be low in protein and oil as well as in weight per thousand 

 seeds. 



Tabulated data relating to the Xenia characters of 20 crosses of Chinese corn, 

 which appeared for the most part to follow Mendel's law, ajid a historical 

 account of maize in China are also presenteiL 



Experiments with corn, E. G. Montgomery {NehmsJca Sta. Bui. 112, pp. 5- 

 36, figfi. 11). — Two ear-row methods of corn breeding are outlined, and the 

 results of each stated. In the first method, taken up at this station in 1903, 

 101 ears of Hogue Yellow Dent were used as foundation stock. If a family (all 

 rows tracing to an original mother plant) did not average well, all of its seed 

 was discarded regardless of high yields of one or more rows. No attempt was 

 made to prevent intercrossing with the poorer families planted near the better 

 ones. During 1907 and 1908 selected strains improved by this method gave 

 yields averaging 8.3 bu. per acre more than did the original stock. 



The second method, instituted in 1906 with 204 ears of the same variety, 

 produced as good results in much less time and by a simpler process. One-third 

 of each ear was planted in a row 16 rods long. The renniants of these ears 

 were retained and in 1907 arranged in order according to the yield from the 

 seed taken from them in 1906 and planted in this order, those making the 

 highest yield being thus grouped together. " The progeny of the four best ears 

 yielded 9 bu. per acre better than the checks, while the remnants of the original 

 ears yielded 11 bu. better." 



The physical characters of the corn plant that are to be considered in seed 

 selection are discussed in full, and directions for conducting breeding plats for 

 a period of 3 years are presented, together with forms for keeping the notes on 

 original seed ears and on the ci'op from those ears. A type of Reid Yellow 

 Dent, averaging 0.2 oz. less per ear in weight. 0.4 in. less in circumference, 0.7 

 in. more in length, and 4 in. less in total length per 100 grains than the standard 

 type, is designated as a long smooth type of Held Yellow Dent, and produced a 

 greater yield than the standard type by 4.4 bu. per acre, with a shrinkage lower 

 by 0.6 per cent. 



Data on the relation of yield and size of ear during 1906 showed that the 

 greatest number of bushels was produced with ears averaging 9.12 oz., while 

 in 1907 the average weight per ear from the plats making the highest yield 

 was 9.76 oz. Phenomena observed in the high-yielding rows were (1) low 

 number of barren plants, (2) large number of 2-eared plants, and (3) large 

 number of tillers bearing ears. Investigations failed to show any definite 

 relation between leafiness of plants or strains and their productive capacity. 

 Data indicate that from 1.732 to 3.627 sq. in. of leaf area were required in the 

 production of a gram of dry matter. 



The work reported on the economic value of tillers is a continuation of that 

 previously noted (E. S. R., 17, p. 657) and in general confirms the results then 

 reported. 



Very little difference in yield resulted from the different methods of dis- 

 tributing seed regularly in hills or irregularly, as in drills. As compared with 

 1 plant per hill. 3 and 5 plants per hill produced increased yields of 0.4 bu. and 

 2.6 bu. per acre, respectively. 



