34 
though we are not yet certain. Dr. ]\IcKenny and myself having 
joined forces to settle, if possible, the problems relating to banana 
diseases in these regions. Possibly there are two banana dis- 
eases now confused — one due to bacteria, the other to fungi. 
A microscopic examination of the Cuban material showed bac- 
teria to be present in some of the vessels but not in quantity 
sufficient to lead me to suppose them to be the cause of the dis- 
ease. In passing, I might say that Earle sent me cultures of 
the bacteria isolated by him from the diseased Jamaican bananas, 
and that in the summer of 1904 I inoculated these copiously into 
the leaf-blades and petioles of bananas in Washington, but with- 
out production of any disease. In the Cuban plants no fungi 
were observed at first, but further studies revealed a small amount 
of mycelium running in the vessel walls or their vicinity, but in 
no case plugging the lumen of the vessels. No spores were ob- 
served at first, but after a while I thought I made out, although 
rather indistinctly, one or two microconidia and jumped to the 
conclusion that the fungus was a Fitsarium. Poured-plates were 
then made from the interior of afifected leaf-stalks which were 
sound on the surface and a Fusariiim was obtained on the plates 
in practically pure culture, the colonies having evidently been 
derived from microconidia present in the bundles. Transfers 
were made from these colonies, and after a half-year or more, 
rapidly growing large banana trees were inoculated from sub- 
cultures. The inoculations were made by means of punctures 
into the midrib, leaf-stalk and pseudo-trunk. At this time the 
bananas were about twenty feet high, perfectly healthy and with 
trunks a foot in diameter. As a result of these inoculations the 
writer obtained infection of the vascular bundles of the petiole 
of several leaves to a distance of from five to eight feet and 
more from the point of inoculation. The bundles became brown- 
purple in the typical manner, and the Fiisarium with micro- 
conidia was demonstrated in the interior of these bundles by 
microscopic examination, especially after treatment with 10 per 
cent, potash, and was also isolated from the same at this distance 
from the point of inoculation by means of Petri-dish poured- 
plates, the exterior of these petioles being at the time perfectly 
sound. It has thus been demonstrated beyond dispute that the 
afifected Cuban plants contain a Fitsarium which is able to run 
long distances inside of the vascular bundles, and cause a purple, 
purple-brown or blackish stain of the same. What has not been 
demonstrated is that such inoculations will so disease the root- 
stock that other uninoculated leaves will subsequently show the 
typical signs of the disease, I was obliged to break ofif this ex- 
periment after about two months, owing to the necessity of mov- 
ing the hothouse and building another one before experi- 
ments could be continued. The root-stocks from which the in- 
oculated infected leaves were cut away have, however, been 
planted out in the new house, and additional inoculations have 
