2 Journal of Agricultural Research voi. iii, No. i 



satisfactory for sealing the openings about the stems consists of a mixture 

 of four parts of unrefined beeswax with one part of tallow. 



Six pots of plants of each variety ' were used, and the water require- 

 ment of each pot was determined independently, in order to provide a 

 basis for the calculation of the probable error of the mean. In making 

 this calculation, Peter's abridged method, based upon the sum of the 

 departures, has been employed. - 



The term "water requirement," when employed in the following pages 

 without further restriction, indicates the ratio of the weight of water 

 absorbed by a plant during its growth to the weight of the dry matter 

 produced, exclusive of the roots (Briggs and Shantz, 1913a, p. 7). 

 The plants were dried to constant weight in a steam-heated oven, main- 

 tained at approximately 110° C. When the plants produced grain, the 

 water requirement based upon the weight of the dry grain is also given. 

 The percentage of grain produced by the plants grown in the pots usually 

 compared favorably with the field performance. Unless a normal per- 

 centage of grain is produced, the water-requirement ratio based on grain 

 production should not be applied to crops grown under field conditions. 

 In a few instances the water requirement based upon the weight of roots 

 or tubers has also been determined. 



SCREENED INCLOSURE 



To protect the plants from birds and severe hail and wind storms, it 

 was found necessary to conduct the experiments in a screened inclo- 

 sure. The inclosure used in 191 1 consisted of a wooden framework 

 covered with wire netting of X-'nch mesh. This framework shaded the 

 plants somewhat, being made sufficiently rigid to support a track above 

 each row of cans, from which the cans were suspended during weighing. 

 To reduce the shading effect, a new inclosure was provided in 191 2, the 

 framework of which was made of i-inch galvanized-iron pipe with pipe 

 posts 9 feet high at intervals of 8 feet. The framework to a height of 3 

 feet was covered with a wooden wall which came slightly above the top 

 of the pots. The remainder was covered with No. 21 galvanized-wire 

 netting of 3/^-inch mesh. General views of the inclosure are shown in 

 Plate I. 



Although the new inclosure reduced the shading effect, pyrheliometric 

 and total radiation measurements made inside and outside the inclosure 

 still showed a measureable reduction in the radiation due to the shade 

 of the screen. Measurements made with an Abbot silver-disk pyrheli- 

 ometer (Abbot, 191 1) showed that the intensity of the direct radiation 



' The recorded strains used in these measurements were obtained from the following offices of the Bureau 

 of Plant Industry: Foreign Seed and Plant Introduction (S. P. I.); Cereal Investigations (C. I.); Alkali 

 and Drought Resistant Plant Investigations (A. D. I.). 



2 The formula used was RTn=o.84S ' — , , where Rnj=the probable error of the mean. Id=the sum of 



n\n~j 



the departures, and «=number of determinations. 



A probable error based upon six determinations does not necessarily represent strictly the actual fre- 

 quency diagram, and this must be borne in mind in the consideration of probable errors. For a discussion 

 of the probable error when the number of obser\'ations is small, see "Student" (1908). 



