438 EXPERIMENT STATION RECORD. 



detection by H. F. Roberts; the relatiou of bacteria to alfalfa by W. E. King; 

 the composition and digestibility of alfalfa and its relation to soil fertility by 

 J. T. Willard ; alfalfa feeding by li. J. Kinzer and G. C. Wheeler ; alfalfa for a 

 dairy farm by J. C. Kendall ; and insects and other animals injurious to alfalfa 

 by T. J. Headlee. In these papers the history, culture, improvement, and uses 

 of alfalfa are discussed. The Turkestan, German. American, Arabian, and 

 Peruvian types are noted, the inoculation of soils for alfalfa culture described, 

 and the composition and digestibility of alfalfa considered. The uses of alfalfa 

 for the different liinds of live stocl^ are pointed out, and the methods of im- 

 proving the plant by breeding and selection are discussed. A number of the 

 papers report experimental results. 



On June IS, 190.3, 1,183 alfalfa plants were counted in a plat seeded the fall 

 before and containing 100 sq. ft. On June 20. 1906, this field showed only 670 

 plants and on August 9, 1907, 408 plants. The decrease in the number of plants 

 during a period of HO months was 730, or a loss of 64,4 per cent. Another plat 

 seeded in the spring of 1904 contained 1,130 plants on June 20. on an area of 

 5 ft. square. May 1, 1905, the number of plants had been reduced to 403, and 

 May 25, 1908, to 14S plants. Only 13 per cent of the plants counted in the 

 original seeding survived 4 years after seeding. Although the number of plants 

 was greatly reduced in these plats the stand of alfalfa was apparently as good 

 as ever, because the plants which remained increased in size and sent out more 

 shoots or stems. 



Disliing alfalfa with the conmion disk harrow iu 1903, 1904, and 1905, gave 

 an average yield of 9,922 lbs. of cured hay, as compared with 10,209 lbs. where 

 no disking was done. An experiment in manuring alfalfa at the rate of about 

 10 tons of well-rotted barnyard manure iier acre, applied to the winter wheat 

 crop immediately preceding the alfalfa, produced 63.7 plants per square foot 

 averaging 13 iu. high, while on similar land uot manured 50.4 plants, 4^ in. high 

 were produced per square foot. 



In germination tests at the station the average germination from brown, 

 blackened, and green immature seed in 21 samples was 65 per cent, the range 

 lying between 39 and 80 per cent. The average germination of 5 high grade 

 samples was 98 i»er cent. The api)arently bad seed from these high grade sam- 

 ples gave an average germination of 70 per cent, while an average from 4 low 

 grade samples gave a germination of 75 per cent and 46 i)er cent for the whole 

 sample and for the apparently bad seed, respectively. In one sample the dead 

 and defective seed reached 60 per cent of the total, from 5 to 10 per cent being 

 very common in the samples under test. One of the best samples analyzed 95.6 

 l)er cent pure, while a germination test showed but 63.9 per cent of pure good 

 alfalfa seed, the impurities amounting to 36.1 per cent and consisting of dead 

 and decayed seed 28.1 iier cent, foreign seed 7.1, and inert matter 0.9. Attention 

 is called to the fact that a pound of this sample would have contained 1.35,053 

 good alfalfa seed capable of germination, .59.389 dead and defective alfalfa seed 

 incapable of germinating, and 20,5J)8 foreign seeds. 



In studying the analyses of alfalfa seed made at the station it was found 

 that about 97 per cent contained fi'om zero to 10 per cent of both inert matter 

 and foreign seed. While 60 per cent of the samples analyzed contained from 

 zero to 10 per cent of foreign seed, 50 per cent of these samples contained from 

 zero to 10 per cent of inert matter. In 35.2 per cent of the samples subjected 

 to analysis the seed failing to germinate ranged from 15 to 25 per cent, and a 

 germination percentage of not more than 64 was found in 30 iier cent of the 

 total number of samples, while the mean germination percentage of all of the 

 253 samples was 75,4 per cent, A table is given showing that with the poorest 

 10 samples the cost of seeding 1 acre ranged from $3.75 to $34, while the cost 



