spring. Also it rained within 24 hours after the seed was planted and the entire 

 season was rainy. 



The formulas used in this experiment require a few words of exiDlanation. Since 

 there was considerable smut in the treated rows during 1919 it was thought best to 

 cut down the distance treated with one pint of the commercial formaldehyde from 

 3200 to 3000 feet. This also had the advantage of being approximately the same 

 rate of application as was being used in the other states; therefore, our results could 

 be more readily compared. The formulas 1-48-3000 and 1-32-3000 were added to 

 see whether a more concentrated solution would give better control without injury 

 under field conditions. In order to test the effect of increasing the length of row 

 treated with one pint of commercial formaldehyde, the formulas 1-50-4000 and 

 1-50-5000 were added. From the standpoint of the growers, the 1-50 dilution is 

 the most convenient because most of them use 50 gallon barrels for drawing the 

 water and they all have gallon jugs or cans. It is, therefore, a simple matter to add 

 a gallon of the commercial solution to each barrel of water. 



Observations on the course of the disease throughout the summer coincided with 

 those of the previous season. The percentage of smutted plants was not counted 

 during the growing season for reasons previously stated. All the onions were pulled 

 and counted on September 4. A week later they were topped and weighed. The 

 data taken are recorded in Table II (p. 27). 



In interpreting the results of this experiment the fact that the soil was very wet 

 at the time of planting must be taken into consideration. The important influence 

 which this factor exerted was understood only after the experiments of later years. 

 We may draw the following conclusions from the data presented in Table II: 



1. The highest percentage of control was secured by the use of the concentrated 

 solutions: thus, the 1-64-3000 was the best*, but not much better than 1-32-3000 

 and 1-96-3000. 



2. No advantage was gained by the extreme dilution and use of the large amount 

 of water required ifor the 1-128-3000 formula. Three of the other formulas gave 

 better control and others were not much inferior. 



3. The decided drop in the percentage of control secured by the 1-50-5000 

 formula indicates that (during a wet spring) the amount of formaldehyde was in- 

 sufficient. It resulted in a fair amount of control — 40 per cent, increase over the 

 untreated plots — but much below the benefit secured by the use of more formalde- 

 hyde per acre. The 1-50-4000 formula gave good control, but the yield was not 

 quite so high as for some of the others. 



The poor showing made by the 1-48-3000 formula is unexplainable in view of 

 the good showing of the formulas just above and just below it. 



The Kuzmeski Farm Experiment, 1920 



The field on which this experiment was located is in Leverett and was seriously 

 infested only in spots. The seed was planted May 3, the soil moisture conditions 

 being about the same as pre^dously noted for the Clark farm. Only two formulas 

 were used, viz., the 1-96-3000 formula which approximated the formula which gave 

 the best control during the preceding season, and the 1-50-5000 formula previously 

 mentioned in the account of the C. A. Clark farm experiment. Eight rows, each 

 436 feet lopg, 13 inches apart, were treated with each formula, and four rows were 

 left untreated as a check. Data for this experiment are recorded in Table III (p.27). 



The results shown in the table confirm the conclusion from the Clark farm ex- 

 periment that the use of one pint of formaldehyde to 5000 feet of drill is insufficient 

 for the control of smut during a wet spring. Entirely satisfactory control was ob- 

 tained by the use of the 1-96-3000 formula. 



C. A. Clark Farm Experiment, 1921 



The experiment of 1920 on this field was practically duplicated during the season 

 of 1921 but with the addition of some more concentrated formulas. The soil, how- 



*In tnis connection, it may be noted that Walker (48:324) during two years out of his four years' ex- 

 periments secured nis highest yields by the use of the 1-64 dilution, while during the third year, it was 

 such a close second to the 1-128 dilution that the difference was almost negligible. Despite these results, 

 he recommended the 1-128 dilution. 



16 



