203 



coiapact to wry loose. 3. 71/)ie of soirh>{/. —Seed was SQ-vvn at inter- 

 vals of one Aveek fi-oni April to 27, 1888, and from March 14 to 

 April 25. ISSl). -i. /)('/>f/i of sowing. — The depths varied from 1 to 

 C. inches. The iiiiinlxM- of plants growing at dirterent dates, yield of 

 grain and straw, and the nnniber of shoots and heads are given in 

 tables. .-). Test of roriefies. — The tests included twenty-nine varieties 

 obtained from seedsmen, and " two varieties, AWlcome and common 

 mixed oats, grown some years with good results on the University 

 farms." These varieties aiv classified as very early, early, niedinm, 

 late, and very late maturing, and each is bi-ielly described. There 

 aie also notes on the vitality and purity of the seed, the yield, date 

 of rii)ening. ]dumpn(>ss of berry, color, kind of panicles, weight of 

 grain per bnsht'l. weight of berries, and quality as indicated by the 

 i-;;tio of kernel to l)erry and the relationshi}) between per cent of 

 kernel in the berry and the weight per bushel, the color, weight, and 

 ])lumjjness of the i)eri-y. together with tabular data. 

 The results from all the experiments reported are thus summarized : 



n) In 18S8 tlie largest yield was produced when 2^ bushels i)er acre were 

 sown; in 1889, when 3^ bushels were sown. Both years considered, the yield 

 was related to the quantity of seed sown from most to least as follows : 3J, 4, 

 3. 2h 2, 1^, and L 



(2) A medium loose seed bed gave a larger yield in both seasons than a com- 

 I)act or very loose seed bed. 



(3) Almost without exception the earlier the seeding the larger the yield 

 ami the greater the weight per bushel. Oats sown March 14. 1880. yielded half 

 and nearly twiee as nui«-h more than oats sown April 4 and 18. respectively. 



(4) Between 1 and 4 inches in depth the differences in yield do not indicate 

 with certainty th.-it the depth of sowing affected the results. 



(5) In many cases varieties of oats under distinct names resemble each 

 other so closely as to be practically identical. 



(C>) Thirty-tlu-ee plats, including varieties under thirty names, gave the rather 

 low average of 41 bushels per acre. The largest yield was 54 and the least 30 

 bushels per acre. 



(7) There was an average of sonn-wliat less than 2 p.)unds (1.S4 pounds) of 

 straw for each pound of gr;iin. 'I'lie variation in tlie yield of straw was a little 

 less than that of the grain. 



(8) The following varieties gave the largest yields, in the order named: 

 Giant Yellow French. Karly Dakota White. Improved American. Japan. White 

 Bonanza, and American ISanner; while Canadian Black. Virginia Winter. White 

 Belgian, Black Tartarian, .-ind Texas Kust-Proof gave the ])oorest yields. 



(9) All things considered, it may i>r.)l)al>ly be fairly concluded tliat the earlier 

 ripening varieties were the more desiralile. 



(10) Neither the length. ])lumpness. or weight of b(>rry. nor the weight per 

 bushel appreciablv influenced the yield. 



(11) The white varieties were considerably superior in yield and weight per 

 bushel to the black and dun-colored varieties. 



(12) The varieties with closed panicles yielded somewhat better than those 

 with open panicles. 



( 13) The average i)er cent of kernel in tl)e berries as sown was (»!».() per cent, 

 and c,."». 1 per cent in tiie cro]). The largest individual difference between two 

 Aarieties w;is 111.:', [ler cent in the seed and 12.7 in the crop. This extreme dif- 



