24 BULLETIN 40, U. S. DEPARTMENT OF AGRICULTURE. 



8 c. c. virus+1 c. c. etlier prepared December 9 (ether again added Decem- 

 ber 26). 

 5 c. c. virus+20 c. c. glycerin prepared December 9. 



5 c. c. Yirus+3 c. c. toluene prepared January 4. 



6 c. c. virus +6 c. c. glycerin prepared January 4. 



The original bottled sap used in these preparations and allowed to 

 undergo free fermentation was also infectious on March 5. Earlier 

 experiments have shown that bottled mosaic sap alone, after under- 

 going fermentation, is able 4 or 5 months later to produce infection. 

 Tests with fermented bottled virus kept for 15 months did not pro- 

 duce infection. 



It is evident that the active causative agents in the sap of mosaic 

 plants show a high degree of resistance to the ordinary destructive 

 agencies. 



ROOT INOCULATION WITH THE DISEASE. 



The writers experiments show conclusively that the mosaic dis- 

 ease may be communicated to healthy plants by direct inoculation 

 of the virus into the roots. Several methods of root inoculation have 

 been tested. In some instances young plants have been pulled up and 

 their roots dipped into the sap of mosaic plants. The treated plants 

 were then reset. In other instances, the plants were pulled up and 

 their roots dusted with dried finely ground mosaic material before 

 resetting. Other plants when very young were transplanted into 

 healthy soil, and when these plants had become well established the 

 soil was removed from some of the roots and mosaic sap poured upon 

 the broken surfaces from a pipette. 



In whatever manner the roots were treated with the virus, the 

 disease usually developed in the plants after 10 or 12 days. The 

 roots of a plant pulled forcibly from the ground are always more or 

 less broken, so that the virus in most instances probably came 

 directly in contact with the broken ends. Root inoculation, how- 

 ever, does not seem to be as certain to produce the mosaic disease in 

 plants as inoculation through the leaf tissues. 



SOIL INFECTION WITH THE DISEASE. 



All experimental data at hand indicate that the occurrence of the 

 mosaic disease in the seed bed is not so closely associated with soil 

 infection as has been supposed. 



The following method has been used to test the effects of germi- 

 nating and growing tobacco plants in badly infected soil material : 

 Square, tightly made frames of wood 10 or 12 inches deep and with- 

 out bottoms were set into a carefully prepared, light, friable soil to 

 a depth of 2 to 3 inches. In some of these frames green mosaic 



