20 MASS. EXPERIMENT STATION BULLETIN 368 



tions of stem taken from seedlings whose leaves had been injured at the time of 

 inoculation. The fungus was not obtained from any of the other trees. 



In an attempt to discover whether or not the causal fungus will persist in an 

 affected tree after all wood showing streaking has been pruned out, tissue plant- 

 ings from the trees used in these experiments have been made in artificial media 

 at intervals extending over a period of two years. The Cephalosporium fungus 

 has not been isolated from any of these trees. 



These results should not be confused with results of pruning following infec- 

 tion taking place in nature. In the latter the problem of latent infection looms 

 large, and fresh untreated cuts or breaks may stimulate re\'ival of the fungus. 



DISCUSSION 



The widespread distribution of a wilt disease of elm and possibly other trees, 

 associated with the genus Cephalosporium, is becoming increasingly evident to 

 those workers interested in the diseases of our principal shade and ornamental 

 trees. The common occurrence of the fungus in diseased elms in Massachusetts 

 indicates the importance of a thorough understanding of the problem which is 

 raised by its very general distribution. Accordingly, an attempt has been made 

 to discover whether or not the disease is specific to the American elm, and the 

 entry and development of the fungus within the host has been studied. 



Reisolations of the causal fungus, accomplished six months after inoculation, 

 from the following elms — American ( Ulmus americana) , English ( U. campestris), 

 Chinese ( U. parvifolia), Siberian ( U. pumila), and the ascendens variety of the 

 American ( U. americana var. ascendetts) — together with the observed entrance 

 of the fungus into their leaves, appears to be sufficient evidence that these trees 

 are all susceptible to the disease. The fact that the fungus does not develop as 

 rapidly in the exotic species of elm would suggest, however, that these varieties 

 are not so seriously affected by the disease as the common American elm and 

 may be considered somewhat resi-stant. 



The results obtained in the inoculations of the Siberian elms ( U. pumila) 

 are directly opposed to the findings of Goss and Frink in Nebraska. These 

 workers report having inoculated eighteen trees of U. pumila by hypodermic 

 injections and by stem and root incisions, with uniformly negative results. In 

 the experiments described here, however, no difficulty was experienced in obtain- 

 ing infection of two trees of this same species at the first attempt. 



Goss and Frink discount the importance of leaves as a court of infection and 

 suggest that the disease is probabh' spread by insects which feed on the young 

 twigs. They found that trees which had not been injured in the leaves or twigs 

 were infected as readily as those which had been wounded. Creager, on the 

 other hand, finds that the leaves are the most common infection court, and 

 that the fungus enters only through wounds. He reports having traced the 

 fungus from the leaf through the vascular strands of the veins, midrib, and 

 petiole into the stem. 



The present exi)eriments confirm in general the findings of Cre-aj^er as opposed 

 to those of Goss and Frink. The fungus has been found (o vnXvr the leaves and 

 twigs when these organs are freshly wounded but not wh(_-n lhc\- have been left 

 intact or an infection court wound allowed to dr\ oul . Il has also been dem- 

 onstrated that the mj'celium will grow in the veins and midrib of the leaf. The 

 progress of the fungus through the leaf into the stem followed very closely the 

 description by Creager. Observations of hundreds of specimens received for 

 laboratory diagnosis show that main- of the elms affected b\- the disease have 



