the formaldehyde without at the same time reducing the actual amount of formal- 

 dehyde per unit of row. 



The amount of injury can be predicted from the moisture condition of the soil 

 on the day of planting. It is not affected by weather conditions during the subse- 

 quent days. 



The grower could save time and labor by changing his formula according to the 

 soil condition at the time of planting. 



More seed should be appUed when formaldehyde is used." 



The final results of the formaldehyde injury investigation were not available 

 before the experiments of 1922 were started. Hence the soil moistm-e and the per- 

 centage of loss from formaldehyde injury were not determined. 



C. A. Clark Farm Experiment, 1922 



The experiment was repeated on the same field during 1922. Twenty-six rows 

 500 feet long, 12 inches apart were sowed at the rate of six pounds of seed per acre 

 on April 26. The soil was dry and no rain fell during the next week which was 

 windy, cool and dry. Different samples of formaldehyde were tried in which the 

 percentage of methyl alcohol was extremely small (4 per cent.) or very high (16 per 

 cent.). Also comparative tests were repeated with the 1-128-3000, 1-64-3000, 

 1-50-5000, and 1-50-4000. Formaldehyde injury was e\ndent when the seed came 

 up but unfortunately no records of its comparative extent or of the soil moistxire 

 conditions were made. The onions were harvested on August 25. The ^deld data 

 are recorded in Table IV (p. 28). It will be noted from this table that the results 

 are quite different from those obtained in 1920 on the same field (cf. Table II). The 

 highest yields were secured with the dilute formula, 1-128-3000, while in 1920 it was 

 the concentrated formulas, 1-32-3000 and 1-64-3000, which gave the highest 

 yields. In 1920, the 1-50-5000 gave the lowest yield while in 1922 it was next to 

 the best. This difference can be easily explained, however, in view of the difference 

 in the moisture conditions of the soil at seeding time. The low yields obtained by 

 the concentrated formulas like 1-50-3000 in 1922 were due to reduction of the stand 

 through formaldehyde injury when the soil was very dry. They do not indicate a 

 lack of control because, when less formaldehyde was used (1-50-5000) the yield was 

 increased. 



As to the effect of the methyl alcohol, the data show a much larger yield through 

 the use of formaldehyde with a low percentage of methyl alcohol. Up to the pres- 

 ent, however, the writers have not had opportunity to repeat this test and would 

 hesitate to draw any conclusions from the results obtained on these two rows during 

 one season. 



Kuzmeski Farm Experiment, 1922 



This was on the same field as was used in 1920. On April 25 when the seed was 

 sowed the soil was very dry and continued so for the next week or more. The exper- 

 iment included 16 rows, 220 feet long, 13 inches apart, seeded at rate of four pounds 

 per acre. The observations throughout the summer coincided with those made on 

 the C. A. Clark farm the same season. Smut was serious only in spots; therefore, 

 the harvest data do not really show how complete was the control (since all the 

 onions in the entire row were counted and therefore considerable stretches of row 

 where there was httle smut to control were included). In the worst places hardly 

 any onions remained standing in the check row while there was a normal number in 

 the treated rows. Only one dilution was used on this field but it was applied to the 

 row at three different rates. The final results (presented in Table V (p. 28) con- 

 firm the conclusions drawn from the C. A. Clark farm experiment of the same year; 

 viz., that the largest yield is obtained by using the 1-50-5000 formula, during a sea- 

 son when the seed must be sowed in a dry soil. The difference is due to formalde- 

 hyde injury and not to lack of control of smut. 



The Objects of the Field Tests of 1923 



Differences in yields secured by the various formulas in 1922 were the results of 

 variation in (1) the control of smut and (2) the number of seeds prevented from 



18 



