384 PLANT GROWTH SUBSTANCES 



incubation period of several days (3). Though the virus generally in- 

 vaded the whole plant the tumors were local in character. Histological 

 studies of virus tumors on sweet clover revealed that they were initiated 

 by tangential divisions in cells of the pericycle, opposite the primary 

 phloem (25). Abnormal cell multiplication rather than cell enlargement 

 was found to be responsible for tumor development. The tumors were 

 composed of a central core of xylem elements consisting of short reticulate 

 tracheids of various widths surrounded by a meristematic zone outside of 

 which was a layer of phloem. Tumor cells were apparently unable to 

 differentiate into the most specialized cell type found in the xylem, 

 the vessel, or into its counterpart in the phloem, the sieve tube. Non- 

 granular, nonvacuolated, smooth, spherical bodies staining intensely 

 with safranine were found in the cytoplasm of some of the tumor cells. 

 Tumor tissue isolated from sorrel roots was successfully grown on a 

 synthetic medium without auxin (2) and found to double its volume 

 about every three weeks. The tissues in culture did not become organized 

 into stems, roots or leaves. When such tissues were grafted to healthy 

 sorrel plants the stock plants developed the systemic tumor disease, 

 showing that the virus was still present in the tissue cultures. The 

 heredity of the host was found to play an important part in tumor reac- 

 tion. Clones of sweet-clover plants were selected which, when infected 

 with the virus, developed minute, scarcely visible tumors; other clones 

 were selected on which a dense mass of tumors developed. In most 

 instances it could be shown that the overgrowth had developed as the 

 result of a wound. 



The most intensely studied of all plant overgrowths is that produced 

 by the crown-gall organism, Agrobacteriwn tumefaciens. Erwin F. Smith 

 who, together with Townsend, discovered the bacterial etiology of 

 crown gall (38) emphasized the resemblance of this plant disease to 

 animal cancer and went so far as to argue, by analogy, that some 

 microorganism must also be associated with malignant tumors in animals. 

 Careful study of animal cancer tissue failed to reveal any microorganisms 

 associated causally with the cancerous condition, for which reason Smith's 

 contention that crown gall is cancer (40,41) was rejected by most workers 

 in the field of cancer research. 



The experimental production of overgrowths on plants by means 

 of the crown-gall organism is influenced first by the strain of the 

 bacteria, second by the nature of the plant, and third by the manner in 



