EXTENT OF CROSS-IN OCULABILITY. 157 



same way from soft gall ol daisy the hardest of hard gall on daisy. 

 But inoculating with an organism plated from hard gall of the apple 

 into actively growing soft daisy stem a series of galls were produced 

 more resembling the original hard gall of the apple from which the 

 colonies came than any typical soft gall of the daisy (PI. III). On 

 the other hand, as already detailed, starting with an organism from 

 apple hairy-root and inoculating into young apple tree roots one 

 developed galls while the others developed hairy-root. From one 

 of the galls, however, on the tree which developed only galls, an 

 organism was plated out which looked typical for what was inserted, 

 and this when inoculated into healthy sugar beet produced both galls 

 and hairy roots, indicating that crown-gall and hairy-root are only 

 two forms of the same disease. Clustered roots also formed on one 

 gall on Brassica. This hypothesis is further borne out both by the 

 fact that the hairy-root clusters often originate in slow-growing hard 

 galls and by the observation that rootlets frequently appear on peach 

 galls in early stages of their development, but do not persist. The 

 same phenomenon, transitory for the most part, occurs frequently 

 or occasionally in some other galls, i. e., daisy, grape, clover, alfalfa. 

 Attempts at cross-inoculation have shown numerous differences 

 (Tables II and III) the explanation of many of wliich must be sought 

 in further experiments. Strains taken from some hosts, e. g., daisy, 

 peach, hop, were inoculated into other plants with great ease. The 

 strain obtained from the rose was inoculated into other plants with 

 difficulty, but inasmuch as tumors were not readily produced on the 

 rose itself by such inoculations it may be only that we were unfortu- 

 nate in the selection of our rose bushes or of the colonies for our 

 subcultures, getting slightly virulent strams. It is certain from our 

 experiments on the daisy that a virulent strain may gradually lose 

 its power to infect when kept for several years under laboratory con- 

 ditions, and it is very probable that in nature some strains are feebly 

 infectious and others actively infectious. 



But we know in case of certam bacterial organisms cultivated in 

 the laboratory that lessened virulence can be restored by certain 

 procedures and we are not warranted in assuming that such restora- 

 tion may not also take place in tlic fields. 



DISCUSSION OF QUESTION OF SPECIES, VARIETIES, AND 

 RACES OF THE CROWN-GALL ORGANISM. 



Have we to do with one species or several ? The answer is not at 

 hand. Indeed, to those who have read thus far, it must be evident 

 that much further time ^^i^ be required to decide positively whether 

 it is best to regard all croAvai-galls as due to variations of one polymor- 

 phous species, or whether they should be separated into two or more 



213 



