398 REPORT OF THE HARVARD AFRICAN EXPEDITION 
the affection in Liberia shortly after we started into the interior. However, 
we have not noticed it previously in tropical countries where mandioca is com- 
mon, as in Central or South America. Notwithstanding the fact that it is 
so common in parts of Africa, we have found no written description of it and 
nothing seems to be known of its etiology. The Acting Director of the Botani- 
cal Gardens at Coquilhatville in the Belgian Congo, to whom we spoke later 
of the affection, said that while he had of course observed it, its cause was quite 
unknown. 
In the affected plants, many of the leaves are deformed, distorted, curled 
and shrivelled, and frequently have a wilted appearance, while some of the 
plants are very much stunted. Often there are irregular areas in the leaves 
which are of a less dark green or of a yellow color in which there is evidently 
a marked loss of the chlorophyl (Nos. 338-340). The condition of the leaves 
suggested a form of mosaic disease or infectious chlorosis, and that it was 
apparently not a form of heredity or of malnutrition chlorosis. However, it seems 
possible that the condition deseribed under the term ‘‘mosaic disease’ may 
represent a group of infections or at least more than one type of infection, and 
that the term mosaic disease has sometimes been applied more to a symp- 
tomatic condition. 
One of the first mosaic diseases discovered, that of the tobacco plant, is 
generally regarded as being due to a filterable virus, Ivanoski! being the first 
one to show that the affection could be transmitted by the filtered juice of the 
plants. Olitsky ? reported he had cultivated this virus or at least he was able 
to show that the active agent remained alive in sterile juices of the plant for 
thirty-three days and that transfer of it could be made from tube to tube. 
However, Goldsworthy,? Mulvania,* and Purdy *® have been unable to confirm 
Olitsky’s work upon cultivation. Up to the present time no plant virus has 
been proved to multiply except when associated with living plant cells. 
Recently Murphy ° has pointed out that the virus diseases of plants give 
the impression of being a more homogenous group than animal virus diseases. 
He emphasizes that while it is easy to filter the sap of the plant and to obtain 
a filtrate free from any visible life, in numerous diseases of plants, presumed 
to be virus diseases, it is not possible by mechanical processes to infect the 
plant with the filtrate. On the other hand, virus diseases can be transmitted 
from plant to plant by grafting or pruning different plants, and this is looked 
upon as the hall mark of virus disease in plants. He also points out that a char- 
acteristic of virus diseases of plants is the fact that they are systemic and affect 
all the plant tissues, leaves, stem, root, everything except the seed, which, 
among plant diseases in general, is very unusual. He suggests that the agent 
producing these virus diseases is entirely distinct from the fungi, bacteria, and 
1 Tvyanoski: Cent. Bac. (1893) Botan. Beihft. 3. 
2 Olitsky: Jour. Exp. Med. (1925), XLI, 129. 
’ Goldsworthy: Phytopathology (1926), XVI, 873. 
4 Mulvania: Science (1925), LXII, 37. 
5 Purdy: Bot. Gaz. (1926), LX XXI, 210. 
® Murphy: Proc. Royal Soc., Feb. 28, 1929; published in Brit. Med. Jour. (March 9, 1929), p. 448. 
