172 CROWN-GALL OF PLANTS. 



wounds, and because some of these galls were observed to arise from 

 the deeper tissues and to push up the sound superficial tissues (Plate 

 XXX) several days in advance of any actual rupture of the latter. 



When the first cuttings were made from the first galled plants, 

 notes were kept of the behavior of these cuttings, and, of 33 made, 

 18 developed galls mtliin six weeks. Some of the galls were under- 

 ground on the base of the cutting, some were at the surface of the 

 earth, and some were on the upper part of the stem. 



Finally, experimental inoculations into the leaf-traces under the 

 point of insertion of the leaf, caused, first, the appearance of galls 

 on the stem where the needle entered, and subsequently at a dis- 

 tance, internal galls. These internal galls appeared along the line 

 of the punctured leaf-traces in the petiole and on the midrib of the 

 leaf, several centimeters from the primary galls, and gradually 

 ruptured to the surface. The plants selected were sound and there 

 could be no question of the secondary galls having originated from 

 within, and as a result of sorne stimulus due to the lyrimary gall, both 

 because they appeared exactl}^ where it was reasoned out in advance 

 that they should appear, and because they were watched through 

 all stages from the first slight elevation of some portion of the sound 

 midrib until through stress of internal tensions it finally split open, 

 showing the tumor tissue in the bottom of the cleft, which tissue 

 gradually increased m size until it projected far beyond the borders 

 of the crevice as a typical gall. 



These growths developing from within outward must be due to 

 migrations or growths from the primary tumor (of bacteria cer- 

 tainly, of host cells inclosing the bacteria probably), but we have 

 not been able to demonstrate the channel of migration either in 

 unstained or stained sections. Cuts made at various points between 

 the primary and the secondary gall yielded nothing to the microscope, 

 nor did we obtain bacterial colonies on agar poured plates made from 

 such tissues, but this is not surprising considermg the relative 

 scarcity of the bacteria in the galls themselves. In the olive tuber- 

 cle, which is superficially like the crown-gall, there are abscess 

 cavities filled with the parasitic bacteria, and a distinct channel of 

 infection can be traced from the primary tumor to the secondary 

 (metastatic) tumors. This occurs m the wood following the path 

 of certain spiral vessels situated at the imier border of the xylem 

 next to the pith. Here distinct lesions occur. On cross section 

 the path of migration in the stem can be seen with the naked eye 

 in the form of small brown dots (lines on longitudinal section) from 

 which under very favorable circumstances a white bacterial slime 

 may be seen to ooze in minute quantity. Under the microscope 

 this browned area is seen to be occui)ied by bacteria. The vessels and 



213 



