The defense mechanisms, if they may be so called, that are invoked in plants by infectious organisms appear to be relatively simple; the invader may in some cases stimulate the formation of corky tissues or may produce other growth effects. Automatic walling-off of the parasite occurs with hypersensitive varieties, in which invasion of the fungus is accompanied by prompt collapse of tissue. Thus the paradoxical situation exists in which hypersensitivity limits fungus extension to a mere flecking, and the most susceptible plants are the most resistant.

IT MIGHT SEEM simplest to explain disease resistance as an antagonism of the juices of the host cell to the parasitic invader. But antagonistic chemicals of host plants have mostly remained undiscovered except for the catechol and protocatechuic acid in the red- and brown-skinned onions resistant to Colletotrichum circinans and Botrytis allii, as found by K. P. Link and J. C. Walker in their investigations at the University of Wisconsin in 1933. These chemicals are not found in the susceptible white-skinned types. But the chemical antagonism found by Link and Walker did not extend to all fungi, since Aspergillus niger could grow in extracts from the pigmented onions.

Plants, depending on how they are grown, vary in their resistance and in their chemical make-up. Furthermore,it is difficult to determine biochemical compounds, and so far there are no leads to relate any compound or class of compounds to disease resistance. Because of such obstacles, this field of research has remained almost unexplored. In the two decades since the discovery of resistance in pigmented onions, and the determination of the chemicals responsible, no additional information of parallel exactness has been obtained.

There is no accepted evidence of the production of antibodies in plants as a result of invasion by fungi. Localization in fungus attack, absence in plants of any circulating medium comparable to the blood stream, and the apparently simple chemical reactions involved in many of the lethal effects of fungi upon plant cells may explain why infected plants do not produce protective substances comparable to those known in animals.

Even with the bacterial organism, A. tumefaciens, which (on the basis of the type of cell division, presence of tumor strands, and the profound disturbances in cell morphogenesis) Erwin F. Smith considered as producing a plant cancer, it was not possible to demonstrate antibody formation. In Smith's tests, repeated inoculations of an individual plant did not bring about lessening of reaction.

The virus disease curly top furnishes some evidence of the production of protective substances as a result of infection. This effect differs from saturation phenomena known for tobacco ring spot and interference as known to take place between viruses. J. M. Wallace, working with curly top in the Department of Agriculture in 1944, inoculated tobacco and tomato plants by means of beet leafhoppers, Circulifera tenellus, which had fed on infected plants. The inoculated plants soon showed severe symptoms of curly top. Tomato plants did not recover but many tobacco plants did. The leaves from such plants showed only mild symptoms in a few weeks. If healthy shoots of tobacco or tomato were grafted onto the tobacco plants that showed recovery, only mild symptoms were produced on the scions. The virus in the recovered plants was demonstrated to be highly virulent by insect transmission tests, but the symptoms produced were mild. When cuttings were taken from recovered plants of tobacco, or from tomato plants protected by the scion grafts from recovered tobacco plants, and these in turn were grafted onto tomato plants, the disease produced on the tomato was mild. Wallace attributed these results to protective substances against curly top developed in the respective tobacco and tomato plants in the course of infection. Once a tomato plant is protected, serial grafts give only a succession of mild curly top reactions, the scions taken from such plants serving as donors of the protective substance. If protective substances were to be sought, it would be logical to look for them with a systemic disease such as curly top.

We have considered so far entrance of fungi into plants, the mechanism of disease production, and the responses provoked. For the higher grades of parasitism, certain other essential requirements in disease production, namely, conditions for establishment of the organism within the host, are important. Suggestions have been made as to osmotic differentials which must prevail in host-parasite relationships. According to this, the concentration of the cell sap of the parasite must be higher than that of the crop plant, if the parasite is to take water from the host. No regularity in this respect has been found. To explain resistance of Zea to Puccinia sorghi, the hypothesis was advanced that susceptibility is determined by the presence of a relatively large quantity of a nutritive substance that attracts the fungus after penetration and makes possible abundant development of the rust. In resistant host Plants, this hypothetical substance is Present only in very small quantity; hence the fungus dies of starvation, and that, in turn, leads to necrosis of the host tissue. It has been suggested that the slower development of the pathogen in a resistant host, as compared with abundant growth in a susceptible host, may be ascribed to a more favorable nutritive substance ratio in the latter.

In some investigations at the Michigan Agricultural Experiment Station, conducted with L. J. Klotz, I studied selectivity for host plants of certain species of Cercospora. Cercospora apii causes lesions only on celery and closely related umbellifers. It does not attack plants from other families. Diseased celery leaves gave strong nitrite tests, as compared with absence of nitrite in healthy leaves, indicating that breakdown of proteins and other nitrogenous compounds occurred. Now any of the species of Cercospora will grow readily and indiscriminately on cooked plant tissue, but with respect to living tissue they are very choicy. Furthermore, the pathogenic limits apparently tend to disappear as leaf tissue matures and cells are about to die. It does not seem that a parasite such as Cercospora, once it invades the cells, should have difficulty in appropriating water necessary for its growth. It seems probable also that the carbohydrates, such as sucrose, dextrose, or other sugars, and water-soluble substances exist in plants in much the same available state as in synthetic or cooked media. As a result of our investigations, we suggested that the nitrogen phase of fungus nutrition was particularly important and that the protein-dissolving enzymes that a given fungus has may determine its capacity for attacking a given plant species.