404 REPORT OF THE HARVARD AFRICAN EXPEDITION 
In a subsequent paper! he describes the preparation of ultraviolet light photo- 
graphs of seven typical juices of plants infected with virus diseases, using a 
wave length 275 millimicrons, but no formed structures other than those seen 
in corresponding fluids in healthy plants were found. 
Jones? had previously reported the cultivation of a mycetozoon which he 
named Plasmodiophora tabaci from tobacco juices affected with mosaic disease. 
He further found that identical symptoms were produced following the inocu- 
lation of healthy tobacco plants with such cultures. He was able to find the 
plasmodial stage of the protozoan in tissues of tobacco plants displaying similar 
symptoms of the disease, but healthy plants were entirely free of it. 
However, the more recent investigations of Link and Taliaferro,’ in which 
Jones also collaborated, seem definitely to show that Plasmodiophora tabaci can 
be cultivated from both healthy plants and those affected with mosaic, provided 
they are not washed in an antiseptic solution such as mercuric chloride and that 
the inoculation of tobacco plants with cultures containing various stages of 
P. tabaci is followed by mosaic only when the cultures are derived from diseased 
plants in which case a concomitant mosaic virus could be present and carried by 
pollution. They also show that filtrates from diseased plants which were 
infective for healthy plants did not show any P. tabaci. 
In view of the large amount of evidence that the causative agent of a number 
of the mosaic viruses pass through diatomaceous and porcelain filters, it seems 
evident that the infective organism must consist of a very minute stage. It has 
been pointed out that the intracellular bodies referred to could hardly be of 
etiological significance since they are too large to pass through the pores of a 
porcelain filter. However, the infective units or spores of which they are com- 
posed may be very minute. The recent work of Link * and of Holmes °® all em- 
phasize this fact. 
The results of our histological study of the African mandioca disease are dif- 
ferent obviously, from some of those which have been referred to in connection 
with the mosaic disease of tobacco. The chromatinic dot and rod-shaped bodies 
that we have found, while they resemble somewhat those observed by Holmes in 
Hippeastrum mosaic, stain differently from the chondriosomes. They resemble 
also in form the mitochondria which Cowdry and Scott © have recently found 
to be present and illustrated in Plasmodium kochi and Plasmodium praecox, 
through supravital staining with Janus green. The nature of the round and rod- 
shaped bodies in our sections is not entirely clear. One cannot conclude from 
their appearance that they are bacteria, nor can one assume that they are con- 
nected with the cells of the fungi near which they often lie, nor that they are the 
spores of the fungus. It seems possible that they may represent degenerative 
products but their presence in such small numbers would appear to argue against 
1 Holmes: Bot. Gaz. (1928), LXXXVI, 59. 
2 Jones: Bot. Gaz. (1926), LX XXI, 446. 
3 Link and Taliaferro: Bot. Gaz. (1926), LX XXII, 403. 
4 Link: Loc. cit. 
5 Holmes: Bot. Gaz. (1928), LXXXVI, 66. 
® Cowdry and Scott: Arch. Inst. Pasteur de Tunis (1928), XVII, 233. 
