54 PEACH ROSETTE. 



In another way this experiment establishes the same fact as No. 1, 

 i. e., the gradual transmission of a germ or virus of some sort from dis- 

 eased to healthy parts. Clearly something is now affecting the tissues 

 of the north part of the parent tree which was not in them one year ago. 

 Otherwise the 46 peach shoots should now be diseased exactly like that 

 part of the parent tree. 



It is obvious that some years must elapse before this experiment is 

 complete, but enough has been observed already to make it almost certain 

 that the disease was not communicated by these buds. It is the same 

 kind of an experiment as No. 4 of Part I. 



III. CONCLUSIONS. 



(1) The rosette, as now understood, differs from peach yellows in the 

 following particulars : 



(a) The more tufted character and somewhat different appearance of the diseased 

 growths. 



(&) The much greater tendency of these compactly tufted growths to develop ID 

 early spring from winter buds and to appear all over the tree. 



(c) A less tendency to develop sprouts upon the trunk and main limbs. 



(d) The absence of premature fruit. 



(e) The general early fall of leaves and fruit on affected trees, the fruit being small, 

 yellowish green, and more or less shriveled and gummy. 



(/) Gummosis of the roots. 



(#) The occurrence of the disease in plums. 



(ft) The much more speedy destruction of affected trees. 



(2) The disease is virulently contagious (Experiment 1), and it is 

 probable that something might be done toward checking its increase 

 by the prompt destruction of all affected trees. This should be done 

 in early spring, as soon as the disease appears and before the leaves 

 begin to fall. 



(3) The disease may exist for a short time in part of a tree without 

 being in the rest of it (Experiment 2), but it soon involves the entire 

 tree. In other words, it would seem from Experiment 2 that the cause 

 of the disease must enter the tree at some particular point or points 

 and be carried gradually to all parts through the circulation. A com- 

 parison of the November and June examinations in Experiment 1 also 

 confirms this belief. 



(4) As in peach yellows, the admitted fact that neighboring trees are 

 not always the next to take the disease is no argument against its com- 

 municable nature. 



(5) This disease has gained a strong foothold and is on the increase, 

 especially in that part of Georgia known geologically as the Archa3an. 



(6) If Georgia peach-growers would save their orchards and main- 

 tain the successful cultivation of the peach, the necessity for prompt 

 and concerted action appears to be very great. 



