14 THOMAS: INFECTION OF APIUM GRAVEOLENS 
ment of the fungus than the greenhouse conditions. However, 
the fact that a single leaf in the field bears more infections than 
any count obtained on an entire plant in the greenhouse is note- 
worthy. To obtain some statistical evidence of the relation of 
the fertilizer treatments to the size of the spot, two sets of five 
plants each in three-inch pots were treated respectively with 2 
grams of hydrated lime per plant as top dressing, and 1 gram of 
sodium nitrate per plant in solution. The effect upon both the 
number and extent of infection areas is shown in TABLE V. 
The nitrate plants average 284 spots per plant as opposed to 
120 spots per plant on the limed plants and, what is perhaps even 
more significant, the difference in the size of the spots is propor- 
tional, the spots of the nitrate plants averaging 2.64 mm. in 
diameter while those of the limed plants average only 1.06 mm. 
The ratios are as 1: 2.36 and 1: 2.54, respectively. These data 
indicate that the degree of susceptibility is dependent upon the 
interchanges between the host cells and fungus hyphae rather 
than upon the ability or lack of ability of the fungus to penetrate 
the host. 
In a further attempt to produce varying conditions of health 
in the experimental plants, sets of five plants each were watered 
with 50 c.c. of each of the following solutions: 2 per cent sodium 
chloride, 1 per cent magnesium chloride, 1 per cent barium chloride, 
0.1 per cent ferric chloride, 0.1 per cent zinc chloride. ‘These 
were inoculated with controls as in the preceding experiments. 
The results are not sufficiently uniform to be considered significant. 
The health and growth of the plants were not perceptibly altered 
and the amount of infection was fluctuating. The plants treated 
with magnesium, iron and zinc were somewhat lower in total 
counts than the controls, while the barium series gave the highest 
counts of all. It is to be remembered that our lack of knowedge 
in regard to the behavior of these substances in relation to the soil 
and to the selective absorption phenomena exhibited by the roots 
of plants would make any but the most striking results extremely 
difficult of interpretation. 
Bearing upon the question of the relation of fertilizers to 
infection is the following experiment, which was begun with other. 
matters in view. Five five-inch pots were filled about one-third 
