PECAN ROSETTE IN RELATION TO SOIL DEFICIENCIES. 5 
sible by means of heavy applications of stable manure and cotton- 
seed meal and by the use of legumes. 
In the experiments the most economical method of obtaining these 
results was left out of consideration for the time being. If the cor- 
rectness of the principle could be established, the cheapest method 
of arriving at the desired results in actual orchard practice could 
be worked out by later experiments. Therefore, in order to secure 
quick results, much heavier applications of fertilizers were made 
than would be used in commercial practice. 
Three experiments, including about 13 acres of pecans in three 
orchards in southern Georgia, were begun in the fall of 1915. The 
three tracts were divided into plats, with suitable checks. Three 
plats, one in each tract, received stable manure alone at the rate of 
20 tons to the acre; two received stable manure at the same rate, 
and in addition cottonseed meal at the rate of 1 ton to the acre; 
and two received cottonseed meal alone at the rate of 1 ton to the 
acre. The stable manure was applied in the fall, and the three 
tracts were then plowed ; the cottonseed meal, analyzing 6f per cent 
ammonia, was applied in the following April. In the fall of 1914 
lime alone was applied to a separate plat in one of the tracts at the 
rate of 3 tons to the acre of ground limestone (unburned). From 
April until June clean culture was maintained with harrows in all 
the plats, and they were then sowed to cowpeas. These were per- 
mitted to grow until killed by frost, when they were plowed under, 
and then the entire program was repeated except that the cottonseed 
meal was omitted in the second year. The check plats received the 
same cultural treatment, but no fertilizer. This program has been 
followed without material variation for two years. The trees in 
two of the tracts were 6 years old and in the third 10. None had 
ever borne a crop. The most that had ever been obtained from the 
10-y ear-old trees was a few scattered nuts. The soils in the three 
tracts were not well suited to conserve, or, when needed, give up 
moisture. This was largely due to the heavy clay subsoils with in- 
sufficient surface soil. 
RESULTS OF THE EXPERIMENTS. 
The results of the work for the past two seasons on the three ex- 
perimental tracts have been compiled in Table I. The fact that the 
amount of rosette normally fluctuates to some extent makes this the 
fairest way of presenting the. results. Practically no results were 
obtained during the first season. 
In Table I the term "badly rosetted" indicates trees which are 
dying back and " rosetted " those which are yellow and making but 
little growth. Trees showing merely a trace of the disease were 
