i8 



Journal of Agricultural Research 



Vol. Ill, No. I 



ghums in this respect. The differences exhibited by the remaining varie- 

 ties are without significance when the limitations imposed by the probable 

 errors are considered, although the quick-maturing Northwestern Dent 

 and Hopi varieties appear to be slightly more efficient than the others. 

 The detailed data are given in Table XIII. The pollination was not 

 adequate to give representative grain yields. 



Five varieties of corn and one of teosinte were included in the 191 3 

 measurements. The water requirement of each variety, based on the 

 production of dry matter, is as follows : 



Variety of com or teosinte Water requirement 



Indian Flint com 342 ± 5 



Hopi com 3So±8 



Teosinte, Durango 390±ii 



Northwestern Dent com 399i 12 



Bloody Butcher com 405 ± 7 



China White com 41 5 ±4 



The most efficient varieties were the Indian Flint (PI. VI, fig. 5), 

 a small variety grown by the Indians of northern Michigan, and the 

 Hopi, another dwarf Indian variety. Teosinte, Northwestern Dent com, 

 and Bloody Butcher, a local variety of corn grown near Wray, Colo., 

 showed only slight differences. The China White, as in 1912, proved to 

 be the least efficient of all the varieties tested, having a water require- 

 ment 20 per cent above that of the Indian varieties. 



In 1912 the water requirement of the Northwestern Dent com was in 

 practical agreement with that of the Hopi, while in 191 3 the Hopi gave 

 a considerable lower value. The China White required 32 per cent more 

 water in 1913 than in 1912; the Northwestern Dent, 42 per cent; and 

 the Hopi, 23 per cent. 



The water requirement of certain com hybrids was also measured at 

 Akron in 1912 and 1913. The mean water requirement of each strain 

 has been included in the tables in the summary (Tables XXIX to XXXII) . 



Table XIII. — Water requirement of different varieties of corn and teosinte at Akron, 



Colo., in igi2 and igi^ 



