826 



EXPERIMENT STATION RECORD. 



and hence the desirability of fall plowing' for barley. The root devel- 

 opment of wheat in the 8u})watered pot was somewhat better than in 

 the surface-watered pot but it is thought that the 3'ield would have 

 been greater had the crop in the subwatered pot not been injured by 

 blight. Peas developed especially well in the subwatei'ed j)()t. 



On variations in plants -with special reference to the relation 

 between the grain weight and the nitrogen content of barley, W. 

 JoiiAN^sEN {Mtdd. Caii.sherg Lah.^ 1S90. Xo.}^.^ pp. 228-SLj). — Experi- 

 ments are reported with Carter Goldthorp l)arley which were made 

 with the object of producing a strain especially low in nitrogen, but 

 having a large and plump grain, to meet the re(iuirements of a good 

 malting barley. In the fall of 1893, 500 practically perfect heads of 

 this ^ ariety were picked from S different places in a 1)arley field, out 

 of which the best 86 heads were chosen to form the foundation stock 

 for the study. Five kernels on the same side of each head were taken 

 for analysis and their weights and nitrogen contents determined in 

 case of (nich head. The average weight per kernel (' ' grain weight ") for 

 all 86 samples was 62.28 mg., and the average nitrogen content 1.606 per 

 cent. By grouping the results as to the grain weight, a marked increase 

 in the nitrogen content was found with increasing grain weight. 



In 1894 the kernels from a number of the best heads were sown 

 separately and at harvest 185 of the largest and finest heads were 

 analyzed as in the previous year. The average grain weight was 54.24 

 mg., and the average nitrogen content 1.489 per cent. Here again the 

 nitrogen content increased, as a rule, with the weight of kernel, 

 although there were exceptions to this general rule in about 30 per cent 

 of the cases. 



A])out 50 heads from the crop of 1894 served as seed in 1895, selected 

 mostly from the exceptions mentioned above, and this was repeated 

 the two following years. Through systematic selection of heads with 

 heavy kernels and low nitrogen content for 3 generations, a progeny 

 was obtained in the fourth generation, samples of which were char- 

 acterized by a somewhat higher average grain weight and an appre- 

 ciably lowei" nitrogen content than that found in the rest of the crop. 

 The following summary shows the results obtained with 2 samples and 

 their progeny: 



Grain weight and nitrogen, cuntent of barJcij from selected .seed. 



