Nov. i6, 1914) Pecan Rosette 1 59 



and I a doubtful trace of rosette. Counts in the two adjacent nursery 

 rows on either side showed at least as high a percentage of rosette as 

 those worked with diseased buds. 



Buds from badly resetted branches were taken from Tallahassee, Fla., 

 and were worked on 8 healthy orchard trees at Glen St. Mary, Fla., in 

 August, 1 91 2. At the same time healthy buds were inserted in branches 

 of 2 trees with distinct symptoms of rosette. Out of 30 to 35 diseased 

 buds inserted, 14 developed, and the following August no rosette could 

 be found on any of them. Of the 12 healthy buds on rosetted stock, 

 4 had lived. In the case of i bud the shoot was perfectly normal, but 

 the tree as a whole had meanwhile recovered from rosette. The 3 

 others showed only traces of rosette, but the tree on which they were 

 worked also had nearly recovered from the disease. 



Grafting was also attempted in this connection at Washington, D. C. 

 Sixty I -year-old seedlings were grafted by the veneer method with 

 rosetted cions from Cairo, Ga., but none of the cions developed. 



In February, 191 2, 105 badly rosetted cions from two orchards in 

 Tallahassee were whip-grafted into a part of the general nursery of Mr. 

 H. K. Miller, at Monticello, Fla. Nursery trees on all sides were grafted 

 to healthy cions. The following August 45 cions were living. Of these, 

 7 showed traces and 2 had developed distinct symptoms of rosette. 

 Counts in the adjacent general nursery showed about the same percentage 

 of rosette. 



It will be noted that normal buds and cions on rosetted stocks inva- 

 riably gave rosetted shoots. Rosetted buds and cions on apparently 

 healthy stocks, with but few exceptions, gave healthy shoots, and wher- 

 ever exceptions occurred the percentage of rosetted shoots was no 

 greater than in adjacent stocks worked with normal buds. The results 

 here tend to show that pecan rosette is not caused by a perennial my- 

 celium, or by bacteria, or by any infecting virus within the tissues of the 



host. 



FERTILIZER EXPERIMENTS 



A fertilizer test was started in March, 1902, in a badly rosetted or- 

 chard belonging to Mr. J. B. Wight, Cairo, Ga. Alternate rows were used 

 for the five plots, and the intervening rows in each case were left un- 

 treated. Plot I received nitrate of soda; plot 2, lime; plot 3, cottonseed 

 meal, acid phosphate, and kainit; plot 4, a liberal application of stable 

 manure; plot 5, ground bone meal. Observations in the summer of 

 1904 in plot I showed 5 trees with the same amount of rosette as at the 

 beginning of the experiment; i tree, better; and 5, worse. In plot 2 there 

 was no change in 3 trees, but 10 others were worse. In plot 3, 7 trees were 

 in the same condition as at the beginning, i was better, and 3 were worse. 

 In plot 4, 9 trees were the same, i was better, and 3 were worse. In plot 

 5, 6 trees were the same, 2 were better, and 4 worse. Two check rows in 

 the same orchard showed, respectively, 6 trees in the same condition, i 



