400 REPORT OF THE HARVARD AFRICAN EXPEDITION 
protozoa which produce othér diseases in plants. Keeping these definitions 
in mind, further experimental work including transmission of the disease by 
filtered juices apparently will be necessary before the African mandioca dis- 
ease to which we are referring can be definitely included in the mosaic or virus 
diseases of plants. 
A number of workers have described amoeba-like bodies in the cells of 
tobacco plants with mosaic, and recently plasma-like bodies have been found in 
the cells of plants affected by several virus diseases. McKinley,' in his admirable 
monograph upon the general subject of filterable virus diseases, has reviewed 
very completely this work, and has discussed the different, bacterial, enzyme, 
filterable virus, and protozoan theories in connection with the etiology of 
mosaic diseases. He believes that there is abundant evidence to support the 
filterable virus theory, to which view most investigators are inclined. 
Purdy ? has recently given some evidence in favor of the infectious nature 
of these diseases. She has shown by immunologic reactions with tobacco mosaic 
virus, injected intravenously into rabbits, that there is some evidence that a 
specific antibody to tobacco mosaic virus-sap, lytic in nature, is present in the 
homologous antiserum. 
In Liberia and the Belgian Congo we made a microscopical study of the 
different parts of the diseased mandioca plants, the roots, stalk, and leaves. We 
were unable to find any animal parasites present and no forms resembling 
amoeboid protozoa were seen. However, in the latex of the main stalk and 
of the stem at the base of the leaves of a number of diseased plants, both rounded 
uniformly-staining cells from about 2 to 4u in diameter, suggesting forms of a 
fungus, and smaller, slender bacillary bodies up to about 2u in length, have 
been found in some plants in considerable numbers. After studying for some 
time and finding these small bacillary bodies in preparations of the latex of 
diseased plants only (and not in the latex of healthy ones), they were finally 
found in small numbers in healthy plants; therefore the presence of the bacil- 
lary bodies seemed of much less significance. However, it should be mentioned 
that these apparently healthy plants were growing in the same region with 
diseased ones, since it has been pointed out by East, Weston * Allard, and 
other investigators that in sugar cane and tobacco mosaic, the plants may 
nevertheless harbor the virus and be infective although they are apparently 
healthy. In other words, the infection may be dormant within them. Murphy 
has also pointed out that unlike animals infected with virus diseases the ma- 
jority of plants infected are “carriers.” 
Cultures were made from the latex of a number of diseased plants after a 
thorough burning of the surface of the stalks. No growths of bacteria or of the 
bacillary bodies were obtained, but cultures of a fungus were obtained on two 
different occasions in which the colonies always showed a reddish-brown tinge. 
A further investigation was to have been made of this fungus, but after our return 
1 McKinley: Phil. Jour. Science (1929), XX XIX, 344. 
2 Purdy: Jour. Exper. Med. (1929), XLIX, 919. 
’ East and Weston: Contributions, Harvard Inst. Trop. Biol. and Med. (1925), p. 15. 
