In an experiment to evaluate the various aspects of symptom of witches' broom of jujube tree with two different kinds of scions, taken from diseased plant and stem- grafted upon sound stocks-one cut before wintering(stored in cellar), the other after wintering, it was observed that the disease rate of those cut before wintering(Nov.) was whereas those cut after wintering(March) showed much lower rate, initially and gradually increased up to . In another experiment of diseased bud grafted into healthy seedling, the finding made in the following year was that 14 stocks(only one of which salt union by callusing) were infected out of the given 23- in the initial stage only 4, yet gradually increased to the number of 14. The shoots from the ground portion of the diseased stock were in general more quickly subjected to the disease than the others. Under natural condition, the diseased trees develop at first seemingly the same leaves as healthy ones, it is not until the branches and loaves grow to a considerable degree that the symptom appears. Once appearing, tile disease grows and the symptom continues to appear as late as in Sept. and early Oct., causing tile plant to develop the extraordinary branches and leaves of extremely reduced size, the typical symptom of witches' broom. Such phenomenon can be observed in the experiments of the foregoing paragraphs. And the suckers and roots of diseased plant are bound to be infected, it was found out with no exception. Viewing from the results of the above experiments and observation, it is believed that the low temperature during winter causes the virus in the above-ground portion of diseased plant to diminish or inactivated, and subsequently the virus in the roots moves up or multiply in the sprouting season.