684 EXPERIMENT STATION RECORD. 



An injurious weed, C. French {Jour. Dept. Agr. Victoria, £ (1.903), No 1, pp. 

 65-67). — A description is given of St. John's Wort [Hypericum perforatum), which 

 has become a serious pest in parts of Australia. Various suggestions are given for its 

 enulication and estimates of the proba1)le cost. 



Some introduced species of dodder in Germany, W. Kinzel (Naturu: Ztaclir. 

 Laud- u. Furstw., 1 {1903), No. 5. pp. 177-180). — Notes are given on a number of 

 si)ecies of Cuscuta which have been recently introduced into Germany in clover and 

 other seeds. Among those mentioned are C. planiflora, C. racemosa, ('. lupuUformis, 

 C. (p'onoini, etc. 



Concerning' the germination of dodder seed, W. Kinzel {Naturu: Ztschr. Land- 

 u. Fordic, 1 {1903), No. 3, pp. 104-110). — The results of the study of the germina- 

 tion of a number of species of Cuscuta are given. 



DISEASES OF PLANTS. 



The Granville tobacco wilt, F. L. Stevens and W. G. Sackett {North Carolina 

 Sta. Bui. 188, pp. 79-96, figK. 1.5). — A preliminary accoimt is given of a very destruc- 

 tive disease of tobacco which has made its appearance in Granville County, N. C. 

 Inasmuch as there may be other wilt diseases found to affect tobacco plants, the 

 authors purpose distinguishing this particular form under the name of the Granville 

 wilt, and have stated that a preliminary description of the disease was given in a 

 press bulletin issued by the station August 22. 



The first indication of the disease is seen in the drooping of the leaves, which 

 become soft and flabby as though suffering from want of water. The wilted leaves 

 soon die and finally the whole stalk is destroyed. At the early stages of the dis- 

 ease a cross section of the stem show"S a yellowish discoloration of the woody por- 

 tion. In more advanced stages the wood is found to be filled with black streaks, 

 and these become so abundant in the later stages that the woody tissue is almost 

 entirely discolored. 



The root seems to be the seat of the original infection, and any plant sufficiently 

 advanced with the disease to show symptoms in its foliage will have its roots in an 

 advanced stage of decay. The disease is most conspicuous in the largest roots, but 

 the smallest fibers on close examination are similarly affected. When all the roots 

 of the i)lant are diseased the death of the plant naturally follows very quickly, and 

 in such cases the progress of the disease in the stem is cut off by the death of the 

 plant. If only one root be affected, the disease in the stem will be on the side near- 

 est the affected root. The spread of the disease is much more rapid in a longitudinal 

 than in a transverse direction. 



There are said to be a number of diseases of tobacco which are recognized by 

 tobacco growers, but the Granville wilt may be readily distinguished by the diseased 

 roots and the black streaks in the wood. Corroborative evidence is at hand which 

 seems to show that the disease increases in severity year by year after the first infec- 

 tion in a field. It is further noticed that soil once seriously affected may recover to 

 some extent if tobacco should not be raised on the ground for a number of years. 

 The disease may be carried from an affected field to one lying below it by the flood 

 water which has been contaminated on the upper field. In the region covered by 

 this report tobacco is the chief money crop, and as some fields are almost entirely 

 destroyed, the loss is very great. 



The history of the observation of the disease and its distribution are briefly out- 

 lined, and a number of factors are suggested as to its probable cause. As a result of 

 a careful examination, the authors have found the blackened portions of the root, 

 stem, and leaves to be filled with bacteria. They are never present in healthy parts 

 of the plant, but being so abundant in the diseased portions it is strongly believed 

 that these germs cause the Granville wilt. This hypothesis can not be definitely 



