Nov. i6, 1914] Pecan Rosette 169 



had been reached, trees occasionally regained their normal appearance 

 without any artificial treatment. From the seasonal variations in 

 rosette recorded from year to year under conditions of uniform cultiva- 

 tion and fertilization it seems highly probable that seasonal climatic 

 changes, such as variation in precipitation, may have at least an indi- 

 rect relation to the prevalence of the disease. In large orchards the 

 more or less simultaneous appearance of rosette in patches, together with 

 its usual limitation to these areas, clearly suggests some definite connec- 

 tion with the soil conditions. 



In general, the ash analyses of resetted and healthy material showed 

 slight to highly variable differences, so that very little positive light was 

 thrown upon the problem by this part of the work (see p. 165). In both 

 leaves and twigs from one orchard the magnesium content was much 

 higher in the resetted than in the normal material, but in two other sets 

 of analyses the percentage was nearly the same in both diseased and 

 healthy material. The percentage of phosphorus was greater in the 

 normal leaves of two sets and in the normal twigs of one set. In the 

 other cases the percentage was about the same in both healthy and dis- 

 eased material. The calcium content was greater in the normal leaves 

 and twigs of two sets and considerably less in the remaining set. The per- 

 centage of potassium was greater in all the resetted material, with the 

 exception of one set of resetted and normal twigs, where there was 

 slightly more potassium in the normal. Other differences shown by the 

 analyses are either slight or greatly variable. 



It thus appears from the results of experiments in pruning and cutting 

 back, transplanting tests, fertilizer experiments, results of subsoil dyna- 

 miting, and orchard records that the disease is directly or indirectly 

 caused by some soil relation. On account of their variable character, 

 the ash analyses shed little light on the problem. 



COMPARISONS WITH OTHER DISEASES 



Leaf-hopper injury suggested itself at one time as a possible cause 

 of rosette of pecans. The rather far-reaching effects of the work of 

 this insect were known, and the demonstration of their causal relation to 

 curly-top of beets ' lent further plausibility to this theory. However, 

 extensive observations have failed to disclose any connection between 

 this insect injury and rosette. Leaf-hopper injury has occasionally been 

 seen on pecans, but its symptoms are distinct and it has occurred both 

 in the presence and absence of rosette. The leaves are often yellowed 

 around the margin, somewhat curled, and if attacked while young, their 

 growth is considerably interfered with. But there are not the distinct 

 yellow mottling over the whole blade between the veins, the raised ap- 

 pearance of the latter, and the tendency to reduction of the leaf blade, 

 followed by the dying back of the shoot from the tip. 



'Shaw, Harry B. The curly-top of beets. U.S. Dcpt. of Agr., Bur. Plant Indus., Bui. 181,46 p., jfie., 

 9 pi. 1910, 



