16 BULLETIN 829, U. S. DEPARTMENT OF AGRICULTURE. 
of the disease would be immeasurably complicated if it were to be- 
come prevalent on such omnipresent weeds. Fortunately, however, 
our observations appear to indicate that the grasses other than cane 
become infected only under conditions favorable to the disease and 
in the near vicinity of infect'ed sugar-cane plants. Infected corn, 
for instance, has been seen by the writer only in Porto Eico, where 
it was growing between the rows of diseased cane stubble. Infected 
rice plants were observed there only once, growing just across a 
narrow dirt road from a badly attacked cane field. At Audubon 
Park, La., attacked sorghum was seen in a similar situation, the 
most remote plants being only about 3 rods from the cane, and the 
percentage of attacked plants decreased in an inverse ratio to the 
distance from the cane. The same was true of crab-grass, which 
was abundant in the sorghum field. These observations are en- 
couraging and tend to offset the disconcerting facts discussed above. 
NATURE OF THE DISEASE. 
INFECTION PHENOMENA. 
Sugar-cane mosaic is an infectious chlorosis, similar in many re- 
spects to the mosaic diseases of tobacco, cucumber, bean, tomato, 
and potato. Evidence of its infectious nature exists in hundreds of 
field observations and in the infection of experimental plants under 
controlled conditions. The well-defined epidemic in Porto Eico, in 
which it has been established that the disease started in a small local 
area and gradually spread from this focus of infection, diseased plants 
being confined within the limits of the ever-increasing infested terri- 
tory and not appearing sporadically at remote points, is convincing. 
It leads to the inevitable conclusion that some virus or inoculum is 
responsible for the appearance of new cases and that the only source 
of inoculum is some plant previously infected with the disease. No 
other explanation accounts satisfactorily for the observed facts. 
Climatic conditions were at first suggested, but the epidemic has 
lasted already for a period of years, during which rainfall, tempera- 
ture, sunshine, and the other factors that go to make up climate 
have been normal. The wearing out of soils was regarded as a 
possible cause, but during the steady progress of the disease it gradu- 
ally encroached upon every conceivable type of soil, including the 
richest and most productive in the island. Strong support was 
given to the idea that it was a case of deterioration or the ''running 
out" of varieties, but when it became evident that all varieties 
present in the invaded district were affected, this idea was aban- 
doned. For the same reason the hypothesis that it is a case of bud 
variations, or "sports," seems highly improbable, and when the 
regular progress of the epidemic is borne in mind, radiating outward 
