200 



sively grown and where the dust and surface drainage waters 

 are laden with the corresponding germs, we have an example 

 of the second agency. The uselessness of artificial inocula- 

 tion of alfalfa in the West, where it is so universally grown, 

 is also apparently to be explained in the same way. We have 

 found the cow pea under all natural conditions to be inde- 

 pendent of artificial inoculation in the South, doubtless be- 

 cause of the same agency. 



The third agency is exemplified in the case of bur clover, 

 the burs of which usually contain some of the soil on which 

 they have grown. The writer's experiments indicate that 

 this plant does not need artificial inoculation if the seed is 

 planted without being hulled. Likewise we have found les- 

 pedeza to be independent of inoculation. 



As perhaps illustrative of the change by which certain 

 bacteria adapt themselves to plants on which they would not 

 originally cause tubercles, we may refer to the fact that on 

 land where vetch and clover during the first year develop few 

 or no tubercles, after a few years of continuous growth of the 

 same plant on the same land, tubercles are found in abund- 

 ance. A case of this kind occurred here ; hairy vetch, an 

 anrbual plant, made a poor growth the first year, a fair growth 

 the second year on the same plot, and a luxuriant develop- 

 ment in subsequent years ; and this, too, in spite of the fact 

 that in the earlier years better seasons occurred and fertili- 

 zation was heavier than in the later years. This particular 

 case may also owe something to the agency of germs trans- 

 ported from an adjacent field where a closely related plant 

 had been grown. 



Let us admit that if grown continuously on the same 

 land for a suflicient length of time, clover and vetch may 

 reach the point of producing a normal supply of tubercles. 

 Can the farmer living in a region where the appropriate root- 

 tubercle bacteria are not abundant afford to wait ou slow- act- 

 ing natural agencies to inoculate his fields? Under such cir- 

 cumstances artificial inoculation must Ve regarded, not as in 

 opposition to natural agencies, but as a means of hastening 

 and increasing their activity. 



