In 1902 several plots of cowpeas were seeded on the soil experiment field at the university. One of those plots (404) had become thoroughly infected with the cowpea bacteria because of its being so situated that more or less surface drainage water flowed over it from an adjacent field upon which cowpeas had been grown for three successive years. Another plot (408), owing to a slightly different situation, had not become infected. The two plots were seeded in July after a crop of oats had been removed from the land. Within three weeks after seeding, numerous root tubercles could be found on the plants on the infected plot. Later on, ten average consecutive plants were taken up as completely as possible, and 412 tubercles were found on the roots, making an average of more than 40 tubercles to the plant. On Plot 408 only an occasional plant was found infected, and such plants would usually have only a single large tubercle on their roots. Ten average plants not infected were collected from Plot 408 for comparison with the ten infected plants from Plot 404. [Illustration: PLATE 3. COWPEAS; EFFECT OF BACTERIA IN ORDINARY ILLINOIS BLACK PRAIRIE SOIL.] Plate 3 shows these two bunches of plants, the infected plants with root tubercles on the _right_, and the plants without tubercles on the _left_. Four more sets of ten plants each were then collected, two sets from Plot 404 and two from Plot 408. Each set of infected plants was separated into three parts, (1) tops, (2) roots, (3) tubercles; and each set of plants not infected was separated into (1) tops, and (2) roots. All of these samples were dried and analyzed for nitrogen. The results obtained are shown in Table 1.