192 STATE HORTICULTURAL SOCIETY. 



Hypothesis Rendered Probable. 



(1) The whole tree is affected when symptoms appear in any part of it. 

 This inference rests upon Experiment 2; a small part of Experiments 4 

 (?) and 5; and almost all of the excisions, fifty-two in number. It is 

 hoped that experiments now under way will settle this point definitely. 



(2) In some cases, perhaps in many, the period of incubation, i. e., the 

 time between the insertion of a diseased bud and the appearance of the 

 disease, is longer than any yet clearly established. This is suggested by 

 many things, especially by the behavior of some of the Barnard trees 

 (Experiment 4); by the Excisions E 3, G 3, and the Wilson tree; and by 

 Professor Goff's tree. 



(3) The disease is also communicated to budded trees in some other 

 way than by bud inoculation. This is probable in case of many young 

 trees, and is almost certain in case of old trees. In some of my experi- 

 ments the period of incubation was very considerable, and this is one of 

 the striking peculiarities of yellows, but it is too brief to account for the 

 appearance of the disease after the second or third year, except possibly 

 in case of some of the trees budded at Colonel Barnard's. The evidence 

 in favor of outside infection becomes stronger in proportion as a tree gets 

 older. Of course, it goes without saying that additional experiments may 

 demonstrate the period of incubation in some cases, and perhaps in many, 

 to be much longer than any yet established. 



(4) The trees are not infected through the blossoms. This is inferred 

 from the result of the excisions, and from the fact that, in some cases, the 

 disease appears to develop between fall and spring, and to stimulate the 

 blossoms themselves to an unnaturally early development. About fifty 

 cases of this kind came under observation during the very mild winter of 

 1889-90, and many additional ones in the spring of 1S91. 



(5) Since diseased trees have been shown to be very full of infectious 

 matter it must be that for unknown reasons much of this fails to find an 

 immediate entrance into healthy trees. Otherwise the peach would soon 

 disappear entirely. 



Inquiries Stiggested by the Foregoing Experiments. 



Three special lines of inquiry are now under consideration, and will 

 receive undivided attention as soon as the laborious experiments with 

 fertilizers have been completed. These are as follows: 



(1) The period of incubation of the disease prior to its first appearance, 

 i. e., the greatest length of time a tree may be affected before it shows any 

 symptoms of yellows. 



(2) The exact nature of the contagium. 



(3) Its method of spread other than by bud inoculation. 



t 



PART II.— PEACH ROSETTE. 

 I. INTRODUCTORY. 



The second part of this report will be devoted to a peculiar disease 

 prevalent in Georgia, and first referred to as probably a southern variety of 

 peach yellows,* but since described under the name of The Peach JRosetie.j 



* Peach Yellows: A preliminary Report, U. S. Dep. Ag. 183S. 



t The Journal of Mycology, Vol. 6, No. iv. . 



